Hey! I'm a CS & Economics undergrad at BITS Pilani ('27), currently in Toronto as a visiting researcher at the DGP Lab (UofT).
I like making computers draw things fast. Mostly, that means writing C++, wrestling with GPUs, and building low-level systems. I also have a knack for backend infrastructure and applied ML when the project calls for it.
I'm actively looking for SDE, Systems Software Engineer, Computer Graphics and Applied Research Engineer roles.
- DGP Lab, UofT: Currently researching realtime Gaussian Splatting workflows. I'm building Houdini plugins with custom BVH structures and engineering sub-50ms ML feature pipelines to make segmentation veery fast.
- Blender (GSoC '25): Spent a summer deep in Blender's C++ codebase. I optimized the production sculpting pipeline for 1M+ vertex meshes using spatial cache reordering (boosting throughput by 35%) and implemented zstd compression for the undo engine to slash memory overhead.
- Vidcentum Technologies: Built async Python/Meilisearch microservices that happily handled 1,000+ req/sec with sub-100ms query latency across 50K+ documents.
- Real-Time Photoelasticity (PG 2026): SIMD-vectorized Monte Carlo ray-marching estimator in C++ and NVIDIA Falcor. Cut per-frame optical integration cost by 4× and eliminated ghosting under dynamic loads.
- procLand: A GPU procedural terrain engine. Uses C++, OpenGL compute shaders, and an optimized SSBO pipeline to push 100K+ polygons in real time from noise.
- Chronofactorem: A Dockerized timetable platform backed by PostgreSQL and async microservices that handles 4,000+ students.
- GaussianSplatRenderUENiagara: Hooking up 3D Gaussian Splatting rendering through custom C++ Niagara Data Interfaces in Unreal Engine for fun.
- LegalTree: Built a RAG pipeline and engineered methods to guide LLMs in pulling structured semantic triplets from 10,000+ legal texts (Paper under review).
I spend most of my time writing C++ and Python, but I'm completely comfortable jumping into Go, Java, or TypeScript when the backend needs it.
- Graphics & Compute: OpenGL, Vulkan, CUDA, Falcor, Houdini HDK, Unreal Engine, GLSL/HLSL, Slang
- Systems & Data: Linux, Docker, PostgreSQL, Redis, CI/CD, Vector Databases
- Environment: NixOS (and yes, I probably spend too much time tweaking my system config).
Always down to talk rendering algorithms, scalable system architecture, or NixOS flakes.